The primary objective of this proposal is to establish objective and quantitative morphological parameters for the subdivision of classical varieties of acute leukemia. Computer-oriented image processing techiques, developed at this institution for chromosome analysis and, more recently, for an automatic white blood cell scanner and analyzer will be adapted for the analysis of leukemic cells. Slides from 40 children with acute lymphoblastic leukemia will be scanned, the images of 100 leukemic cells from each recorded, and the digitized images analyzed using cell size, color and texture parameters developed for leukocyte analysis and identification. The resulting quantitative morphological data will be analyzed for significant correlation with clinical data including clinical response to specific therapy, overall survival and cure expectancy. Once developed, these techniques can be extended to the field of acute myeloblastic leukemia where clinical results are relatively poor and where many different combinations of therapy exist. The overall goal is to develop objective criteria which may help to identify those patients who will (or will not) respond to a given mode of therapy.